AGRICULTURAL WATER MANAGEMENT 1
Type of paper: original research paper (regular paper) 2
3
Title: Effects of saline reclaimed waters and deficit irrigation on Citrus physiology 4
assessed by UAV remote sensing.
5 6
Authors names and addresses:
7
Cristina Romero-Trigueros*1, Pedro A. Nortes1, Juan J. Alarcón1, Johannes E. Hunink2, 8
Margarita Parra1, Sergio Contreras2, Peter Droogers2, Emilio Nicolás1. 9
1Departamento de Riego, Centro de Edafología y Biología Aplicada del Segura, CSIC, P.O. Box 10
164, 30100, Espinardo (Murcia), Spain 11
2Future Water, Paseo Alfonso XIII, 48, 30203, Cartagena, Spain.
12 13
Corresponding author: Cristina Romero-Trigueros 14
Departamento de Riego 15
Centro de Edafología y Biología Aplicada del Segura, CEBAS-CSIC.
16
Campus Espinardo P.O. Box 164, 30100, Espinardo (Murcia), Spain 17
Phone: +34 968 396200 (Ext. 6270). Fax: +34 968 396 213 18
E-mail: [email protected] 19
Number of tables: 7 20
Number of figures: 4 21
Page count: 32 (including this one) 22
Research highlights:
23
• Reclaimed water significantly reduced total chlorophyll in grapefruit and mandarin leaves.
24
• Normalized Difference Vegetation Index (NDVI) was related to gas exchange variations.
25
• Near infrared (NIR) and red (R) domains were the best spectral indicators for both species.
26
• Usefulness of remote sensing for assessing diurnal changes in Citrus physiology was confirmed.
27 28 29
Effects of saline reclaimed waters and deficit irrigation on Citrus physiology assessed by 30
UAV remote sensing.
31
C. Romero-Trigueros1, P.A. Nortes1, J.J. Alarcón1, J.E. Hunink2, M. Parra1, S.
32
Contreras2, P. Droogers2. E. Nicolás1 33
1Departamento de Riego, Centro de Edafología y Biología Aplicada del Segura, CSIC, P.O. Box 164, 34
Campus Universitario de Espinardo, 30100, Espinardo, Murcia, [email protected] 35
2Future Water, Paseo Alfonso XIII, 48, 30203, Cartagena, Spain.
36 37
Abstract 38
The aim was to assess the usefulness of spectral data to detect structural and physiological 39
changes in Citrus crops under water and saline stress. Multispectral images were acquired from 40
a fixed-wing Unmanned Aerial Vehicle (UAV) while concomitant measurements of gas 41
exchange, plant water status, leaf structural traits and chlorophyll were taken in a commercial 42
farm located in southeast Spain with two Citrus species, grapefruit and mandarin irrigated for 43
eight years with saline reclaimed water (RW) combined with regulated deficit irrigation (RDI).
44
Measurements at leaf scale and airborne flights were carried out twice a day, at 7 and 10 GMT.
45
Irrigation with RW decreased gas exchange and leaf dry mass per unit area (LMA) on 46
grapefruit. However, salinity from RW resulted in an increase in pressure potential (ΨP) on 47
mandarin and allowed maintaining net photosynthesis (A) and stomatal conductance (gs) when 48
vapour pressure deficit increased. On both crops, leaf total chlorophyll (Chl T) concentrations 49
were significantly reduced by RW. Moreover, RDI decreased A, gs and stem water potential 50 (Ψs) on grapefruit, independently of water quality. Regarding spectral data, red wavelength (R) 51
was significantly correlated with Chl T (p<0.001), except when mandarin was subjected to 52
stressful climatic conditions (at 10 GMT); since R was influenced, in addition to Chl T, by the 53 plant water and gas exchange status. Near infrared (NIR) was a useful indicator of Ψs, A and gs
54
on both crops. The normalized difference vegetation index (NDVI) was clearly related to gas 55 exchange in both species and to Ψs only on mandarin. Finally, we combined data from both 56
Citrus species and the best indicators were NIR and R. The novelty of this study was to show 57
that diurnal changes in physiological and structural traits of Citrus irrigated with RW combined 58
with RDI can be determined by multispectral images from UAVs.
59
Abbreviations 60
A: Net photosynthesis (µmol·m-2·s-1); AF: Airborne flight; C: Control treatment; Chl T: Total 61
chlorophyll (mg·gFM
-1); Chl a: Chlorophyll a (mg·gFM
-1); Chl b: Chlorophyll b (mg·gFM
-1); EC:
62
Electrical conductivity (dS·m-1); ETc: Crop evapotranspiration (mm·month-1); ETo: Reference 63
evapotranspiration (mm·month-1); GMT: Greenwich Mean Time; gs: Stomatal conductance 64
(mmol·m-2·s-1); LMA: Leaf dry mass per unit area (g·m-2); NDVI: Normalized Difference 65
Vegetation Index; NIR: Near infrared wavelength; ns: Not significant; R: Red wavelength; 66
RDI: regulated deficit irrigation; RS: remote sensing; RW: Reclaimed water; SE: Standard 67
error; TW: Transfer water; t1: Time 1; t2: Time 2; UAV: Unmanned aerial vehicle; VPD:
68
Vapour pressure deficit (KPa); WWTP: Tertiary wastewater treatment plant. Ψs: Steam water 69
potential (MPa); Ψπ: Osmotic potential (MPa); ΨP: Pressure potential (MPa).
70
Keywords: chlorophyll; gas exchange; grapefruit; mandarin; multispectral imagery; precision 71
agriculture; water status. 72
73 74
1. Introduction 75
Irrigation water is not always available (mainly in summer) in the semi-arid 76
Mediterranean areas due to water scarcity (Pedrero et al., 2015). Therefore, irrigation 77
scheduling needs to be precise, and this requires strategies to optimize irrigation water 78
productivity (Tapsuwan et al., 2014). One technique currently in use is the regulated 79
deficit irrigation (RDI) strategy, where water deficits are imposed only during the crop 80
developmental stages that are least sensitive to water stress (Chalmers et al., 1981).
81
Furthermore, current climate change predictions indicate increases in the frequency and 82
intensity of drought periods (Garcia-Galiano et al., 2015; Stocker et al., 2013). In order 83
to overcome this issue, the use of non-conventional water sources such as reclaimed 84
water (RW) (RD 1620/2007) would be an alternative for farmers. On the one hand, RW 85
can be beneficial to crops due to its concentration of macronutrients (N,P,K) (Pedrero et 86
al., 2013); bearing in mind that an excess of them could be lost through leaching and 87
other processes (Romero-Trigueros et al., 2014a). On the other hand, RW may have 88
risks for agriculture because of its high concentration of salts. Therefore, inappropriate 89
management of irrigation with RW can exacerbate problems of secondary salinization 90
and soil degradation at the medium-long term, and finally result in negative impacts on 91
crop physiology, growth, crop quality, etc. (Romero-Trigueros et al., 2014b).
92
In order to be successful, RDI strategies and improved agricultural management need a 93
reliable characterization of the plant water status. This is achieved by measurements at 94
leaf scale, and up-scaling this information to the canopy/field level. Measuring the 95
spectral response of canopies is a non-destructive and rapid method to signal stress early 96
in orchards (Jones and Vaughan, 2010). The acquisition of this information with remote 97
sensing (RS) techniques has proven useful and cost-effective compared to more time- 98
consuming and laborious field techniques based on leaf sampling (González-Dugo et 99
al., 2012).
100
Traditional RS approaches have also a number of drawbacks: satellite imagery often 101
suffers from issues with cloud cover, and remote sensors that are fixed on towers within 102
crop fields are relatively expensive when data from several plots needs to be collected 103
(Anderson and Gaston, 2013). However, in recent years, the use of unmanned airborne 104
vehicles (UAVs) increased thanks to technological advances, cost reductions and the 105
size of sensors. These UAVs could be operated by the farmers themselves to diagnose 106
crop features such as water stress and then adjust their water management practices as 107
needed. Hence, UAV technology can fill the gap of knowledge between the leaf and the 108
canopy by improving both the spatial and the temporal resolution of data on vegetative 109
status (Gago et al., 2015). Nevertheless, the reliability of aerial RS approaches must be 110
assessed with plant-truth data carried out in the field, i.e. with measurements related to 111
plant water status (leaf water potential), gas exchange (net photosynthesis and stomatal 112
conductance), chlorophyll content and leaf structure (Berni et al., 2009b; Contreras et 113
al., 2014; Gago et al., 2013; González-Dugo et al., 2012, 2013; Lelong et al., 2008;
114
Zarco-Tejada et al., 2012).
115
Imagery RS technologies are mainly based on canopies’ wavelength reflectances in the 116
visible, such as red, green and blue, and non-visible range of the spectrum, such as near- 117
infrared (NIR). The remote monitoring of these specific reflectances is commonly 118
performed using visible, multispectral and hyper-spectral cameras (Baluja et al., 2012;
119
Zarco-Tejada et al., 2012, 2013a, 2013b). This reflectance can be used as an indicator of 120
plant status because of its relationship with, among others, leaf pigment composition, 121
plant biophysical or structural parameters and physiological status (Jones and Vaughan, 122
2010). Red wavelengths (R) (660 to 680 nm) specifically are absorbed by leaf 123
chlorophyll (Ollinger, 2011). Because salty environments harm or reduce the 124
functionality and content of chlorophyll in the leaves, reflectance may be proportionally 125
reduced. In the NIR (750 to 1400 nm) domain, the spectral response depends on the 126
multiple scattering of light inside the leaf that is mainly controlled by its internal 127
structure, such as mesophyll thickness and water content (Bonilla et al., 2015).
128
Composite indices integrating data from both domains, such as the Normalized 129
Difference Vegetation Index (NDVI), have shown positive correlations with water 130
stress indicators (water potential and stomatal conductance) in a number of crops (Gago 131
et al., 2015; Glenn et al., 2008). In most cases, the indicators used for this purpose are 132
related to canopy structural changes in different days of the year or growth season, but 133
approaches related with diurnal physiology changes along a single day are rare 134
(Gonzalez-Dugo et al., 2015).
135
In the last years, research focused on checking the different vegetation indices acquired 136
from the UAVs equipped with multi-spectral cameras and then comparing them to field- 137
collected measurements of plant-physiological and structural increased (Berni et al., 138
2009a; Contreras et al., 2014; Lelong et al., 2008; Zarco-Tejada et al., 2013a,b).
139
Drought is one of the most studied stress impulses (Baluja et al., 2012; Gago et al., 140
2015; Pôcas et al., 2015; Rodriguez-Pérez et al., 2007; Stagakis et al., 2012; Zarco- 141
Tejada et al., 2012); however, research on saline stress from RW using UAV technology 142
is limited (Contreras et al., 2014). Besides, studies that evaluate saline and/or water 143
stress tolerances over extended periods are scarce because of the cost and time required 144
for extended periods of time (i.e. multiple years).
145
Salinity stress harms Citrus mainly in two ways: (1) by specific-ion toxicity and (2) by 146
osmotic effects caused by the accumulation of salts. If the stress factor remains, changes 147
in the leaf pigments can arise. In this sense, negative effects of salinity on the 148
chlorophyll content have been reported in Citrus species (Papadakis et al., 2004;
149
Romero-Trigueros et al., 2014b), which constitute one of the most important 150
commercial fruit crops worldwide. The experiment reported on here is the first one to 151
evaluate the diurnal effects of prolonged exposure (eight years) to RW and deficit 152
irrigation on grapefruit and mandarin trees under field conditions by i) measurements of 153
plant water status, gas exchange and chlorophyll in order to obtain the plant-truth data 154
and ii) spectral data, acquired with an UAV, both carried out twice over the course of 155
the day. In addition, the current work sought to assess the usefulness of multispectral 156
imagery to determine the structural and physiological diurnal changes in Citrus crops 157
under water and saline stress.
158
2. Materials and Methods 159
2.1 Site description and irrigation treatments 160
The experiment was conducted in 2015 in a commercial Citrus orchard, located at the 161
northeast of the Region of Murcia in Campotéjar (38º07'18”N, 1°13’15”'W, 132 m 162
above sea level) with a BSk climate by Köppen-Geiger classification (Peel et al., 163
2007).The 1-ha experimental plot was cultivated with i) 11 year-old 'Star Ruby’
164
grapefruit trees (Citrus paradisi Macf) grafted on Macrophylla rootstock [Citrus 165
Macrophylla] planted at 6 x 4 meters and ii) 14 year-old mandarin trees (Citrus 166
clementina cv Orogrande) grafted on Carrizo citrange (Citrus sinensis L. Obs. x 167
Poncirus trifoliate L.) planted at 5 x 3.5 meters. Irrigation was scheduled on the basis of 168
crop evapotranspiration (ETc) accumulated during the previous week. ETc values were 169
estimated by multiplying reference evapotranspiration (ETo), calculated with the 170
Penman-Monteith methodology (Allen et al., 1998), by a monthly local crop coefficient 171
according to Pedrero et al. (2015) for grapefruit and Nicolás et al. (2016) for mandarin.
172
All trees received the same amount of N, P2O5 and K2O through a drip irrigation 173
system: 215-110-150 kg ha–1·year–1 for grapefruit and 215-100-90 kg ha–1·year–1 for 174
mandarin, respectively. Weeds were eradicated in the orchard by applying the farmers’
175
commonly used pest control methods.
176
The experimental plot has been irrigated with two different water sources since 2007. In 177
one case water was pumped from the Tajo-Segura canal (transfer water, TW) and in the 178
other case water was pumped from the North of “Molina de Segura” tertiary wastewater 179
treatment plant (WWTP) (reclaimed water, RW). The latter had high salt and nutrient 180
levels (Table 1) with high electrical conductivity (EC) close to 4 dS·m-1,while for the 181
transfer irrigation water the EC values were close to 1 dS·m-1. Saline water was 182
automatically mixed with water from TW at the irrigation control-head to lower its EC 183
to ≈3 dS·m−1 in order to establish a constant EC during the experiment. This high level 184
of salinity observed in the RW was mainly due to the high concentration of Cl- and Na 185
(Table 1). The boron concentration in RW was considerably higher than that in TW.
186
Moreover, higher concentrations of N, P and K were observed in RW than in TW. The 187
pH was more basic in TW than RW (Table 1). No differences in the concentration of 188
heavy metals were found between the irrigation water sources (data not shown).
189
Two irrigation treatments were established for each water source. The first treatment 190
was a control (C) irrigated throughout the growing season to fully satisfy crop water 191
requirements (100% ETc). The second one was a regulated deficit irrigation (RDI) 192
treatment irrigated similarly to C, except during the second stage of fruit development 193
when it received half the water amount applied to the C (50% ETc). The amount of 194
water applied in 2015 to C was 5945 and 7531 m3·ha-1 for grapefruit and mandarin, 195
respectively, while the water applied to RDI was 4875 and 6175 m3·ha-1 for grapefruit 196
and mandarin, respectively. Therefore, RCD treatments saved about 18% of irrigation 197
water in the case of both species.
198
The experimental design of each irrigation treatment was 4 replicate distributed 199
following a completely randomized design. Each replicate consisted of 12 trees, 200
organized in 3 adjacent rows. Two trees of the middle rows from each replication were 201
used for measurements and the rest acted as guards and were excluded from the study to 202
eliminate potential border effects. A total of 64 trees were used in this study.
203 204 205
2.2 Airborne imagery and image processing 206
A flight campaign was carried out on July 7, 2015 using a fixed-wing UAV (eBee from 207
SenseFly) (Figure 1). Two airborne flights (AFs) were conducted at approximately 100 208
m of altitude over both experimental plots: the first one at 07.00 GMT (t1) and the 209
second at 10.00 GMT (t2). For this study the autopilot was used, following the 210
waypoints of a flight plan created using flight planner software (eMotion). The UAV 211
was mounted with a GPS receiver, altimeter, wind meter and a digital camera that was 212
electronically triggered by the autopilot system to acquire images at the correct 213
positions. The camera used was a Canon IXUS 125 HS digital compact camera that had 214
a 16 megapixel sensor, i.e. 4608 by 3456 pixels, and captured JPEG format images in 215
the green, red and near infrared light range. A total of 110 images per flight were taken 216
and processed into ortho-photos using a Structure from Motion (SfM) workflow 217
(Lucieer et al., 2013) as implemented in the software package Agisoft PhotoScan 218
Professional version 0.9.1.
219
Following previous experiences in the area (Contreras et al., 2014), the spectral data 220
retrieved from the red (R, 600-700 nm) and near-infrared (NIR, 700-900 nm) domains 221
were used to compute the Normalized Difference Vegetation Index (NDVI) as an 222
indicator of the vegetation greenness. Green and dense vegetation has a strong 223
absorption of red light due to the presence of chlorophyll, while cell walls strongly 224
scatter (reflect and transmit) light in the NIR region. NDVI normalizes R and NIR 225
spectral responses in order to provide a combined signal strongly related with the 226
healthy and physiological performance of vegetation (Glenn et al., 2008). Here, NDVI 227
was computed as:
228
) /(
)
(NIR R NIR R
NDVI = − +
229
where NIR and R are the total radiances captured at the top of the sensor and codified as 230
digital numbers in the near-infrared and red domains, respectively. Maps of NDVI 231
values were computed for each experimental plot, and average values were extracted for 232
a buffer circular area of 1m-radius centered at each tree crown in order to minimize the 233
soil background disturbance on the overall spectral response of the crown trees.
234
2.3 Field data collection 235
Physiological and structural measurements at plant scale were conducted on July 7, 236
2015, the same date as UAV flights, and after two weeks of the beginning of deficit 237
irrigation in this season, in order to obtain the plant-truth data. They were carried out 238
twice a day: at 07.00 GMT (t1) and at 10.00 GMT (t2), coinciding with the AFs 239
described in section 2.2.
240
Leaf-scale gas-exchange parameters (net photosynthesis, A, and stomatal conductance, 241
gs) and stem water potential (Ψs) were determined on eight fully-expanded leaves from 242
the mid-shoot area of each tree per treatment (two leaves from each replicate).
243
A and gs were determined with a portable photosynthesis system (LI-6400 Li-Cor, 244
Lincoln, Nebraska, USA) equipped with a clear chamber bottom (6400-08) and a 245
LICOR 6400-01 CO2 injector using a 6 cm2 leaf cuvette. The CO2 concentration in the 246
cuvette was maintained at 400 µmol·mol-1 (≈ambient concentration). Measurements 247
were performed at saturating light intensity (1200 µmol·m-2·s-1) and at ambient air 248
temperature and relative humidity. The air flow was set to 300 mL·min-1. Ψs was 249
measured using a pressure chamber (model 3000; Soil Moisture Equipment Corp., 250
California, USA), according to Scholander et al. (1965), in leaves close to the trunk 251
which had been bagged within foil-covered aluminum envelopes at least 2 h before 252
(Shackel et al., 1997). Leaves from the Ψs measurements at t2 were frozen in liquid 253
nitrogen (-196 ºC) and stored at -30 ºC till analysis. After thawing, osmotic potential 254
(Ψπ) was measured in the extracted sap, according to Gucci et al. (1991), using a 255
WESCOR 5520 vapour pressure osmometer (Wescor Inc., Logan, UT, USA). Pressure 256
potential (ΨP) was calculated as the difference between Ψsand Ψπ. 257
Leaf area was determined using an area meter (LI-3100 Leaf Area Meter, Li-Cor, 258
Lincoln, Nebraska, USA) in twenty leaves per tree collected from the two central trees 259
of each replicate per treatment in the early morning and transported in refrigerated 260
plastic bags to the laboratory. Then, leaves were washed with running tap water 261
followed by rinsing in distilled (Desta, 2014) water and left to drain on a filter paper 262
before being oven dried for at least 2 days at 65 ºC. Later, we determined the dry weight 263
to calculate leaf dry mass per unit area (LMA, g·m-2).
264
Regarding phytotoxic elements, sodium and boron were determined by Inductively 265
Coupled Plasma mass spectrometry (ICP- ICAP 6500 DUO Thermo, Cambridge, UK) 266
and chloride anion by ion chromatography with a Chromatograph Metrohm 267
(Switzerland) in the dried leaves which were ground and digested with a mix of acid 268
nitric (4 mL) and hydrogen peroxide (1 mL).
269
Finally, leaf chlorophyll determination was carried out as described in Romero- 270
Trigueros et al. (2014b).
271
2.4 Statistical analysis 272
A weighted analysis of variance (ANOVA) followed by Tukey ´s test (P≤0.05) were 273
used for assessing differences among treatments. Linear regressions among variables 274
measured in the field and spectral data were calculated. Pearson correlation coefficients 275
were used to assess the significance of these relationships. All statistical analyses were 276
performed using SPSS (vers. 23.0 for Windows, SPSS Inc., Chicago, IL, USA).
277
3. Results and Discussion 278
3.1 Plant water status and leaf structural traits 279
We considered the data presented in this section as truth-plant data because they are 280
field-collected-leaf measurements. Table 2 shows some climate variables for July 7, 281
2015: vapour pressure deficit, mean temperature and average radiance increased from t1
282
to t2, as expected.
283
Plant water status 284
Stem water potential (Ψs) was not influenced by salinity from RW in any of the crops 285
(Figure 2), in agreement with the results found by Nicolás et al. (2016) for mandarin 286
trees. Nevertheless, plant-water relations are proven to be affected by water quality 287
(Paranychianakis et al., 2004). Regarding RDI, there were no significant differences 288
between treatments of grapefruit trees at t1. However, at t2 Ψs of the RDI treatments 289
declined significantly with respect to that of the C treatments: 15% for TW treatments 290
and 11% for RW treatments, as expected. Short-term water deficits may affect plant 291
growth processes and therefore monitoring of water stress is critical not only for early 292
detection of stress, but also for applying RDI strategies (Fereres and Soriano, 2007) 293
with the degree of precision needed. On mandarin trees, the more negative Ψs values at 294
t1 were observed for the C trees for both TW and RW treatments (TW-C and RW-C).
295
This was probably because the well-irrigated trees had at the end of winter 2014 greater 296
plant canopies than the trees under RDI, thus absorbing more water from the soil profile 297
with a consequent lower water potential in the morning. The measurements were carried 298
out only two weeks after the initiation of RDI.
299
On the one hand, both salinity and water stress in grapefruit resulted in a decrease of Ψπ,
300
with a slight increase in ΨP, although in this case no significant differences were 301
observed between treatments (Table 3). On the other hand, in mandarin only the RW 302
treatments (RW-C and RW-RDI) showed a Ψπ more negative than TW treatments and, 303
in this case it resulted in a significant rise in ΨP, similar to findings by Aksoy et al.
304
(1998) and Gimeno et al. (2009) for mandarin and lemon trees, respectively. It is known 305
that when ΨP of ´Carrizo Citrange´ under saline conditions is similar to or higher than 306
that of C trees, Cl- and Na accumulation represent important osmotic adjustment 307
processes and not a significant toxicity effect (Pérez-Pérez et al., 2007). Therefore, 308
according to Aksoy et al. (1998), the response of different Citrus rootstocks under saline 309
conditions is not always similarsince in our case salinity from RW only increased the 310
leaf turgor in mandarin trees and not in grapefruit trees.
311
Gas exchange parameters 312
In the case of grapefruit, both water and saline stress decreased A and gs (Table 4), in 313
agreement with observations by other authors (Anjum, 2008; Hussain et al., 2012;
314
Melgar, 2008). Stomatal conductance in particularly is considered a suitable parameter 315
to assess plant water stress (Flexas et al., 2002). A reduction of this parameter in well- 316
irrigated, but salt-stressed Citrus leaves has also been associated with the specific 317
toxicity of Cl- and/or Na (Levy and Syvertsen, 2004), as probably happened in the case 318
of the RW-C.
319
On mandarin trees at t1, RDI treatments showed A values slightly higher than their 320
corresponding C treatments, but these differences were not significant. This behaviour 321
responded to Ψs (Figure 2). Besides, there was stomatal closure in RW-C with respect to 322
the rest of the treatments (Table 4). In this sense, Ψs regulated physiological processes 323
(Gomes et al., 2004) and induced stomatal closure which reduced A. At t2, unlike with 324
grapefruit, both parameters decreased only in TW-RDI, and not in RW treatments. As 325
mentioned above, one of the main plant adaptations to osmotic stress, e.g. from saline 326
water, is osmotic adjustment which maintains the positive leaf turgor required to keep 327
stomata open and sustain gas exchange (García-Sánchez and Syvertsen, 2006) as 328
occurred in RW treatments. This response has already been described for Citrus, but is 329
rootstock dependent (García-Tejero et al., 2010) since it determines the tolerance or 330
sensitivity to different abiotic stresses, including salinity (Gimeno et al., 2012; Navarro 331
et al., 2011). Our results for example showed that mandarin trees, grafted on ´Carrizo 332
citrange´, increased their ΨP when they were irrigated with RW and, for that reason, gas 333
exchange was unaffected; however, grapefruit trees, grafted on Macrophylla rootstock, 334
responded differently (Table 4).
335
Finally, Citrus trees grown in semi-arid areas are affected by high VPD that induce a 336
continuous decline in gs and A from the early morning hours, even when trees are well- 337
irrigated (Villalobos et al., 2008). In our study, grapefruit trees showed A and gs levels 338
higher than mandarin trees and the lower reduction of both parameters from t1 to t2 was 339
in grapefruit trees: the RW-RDI treatment of grapefruit was the most affected(reduction 340
of 44 and 42% for A and gs, respectively) caused by a water stress and a Na, Cl- and B 341
accumulation (Table 5). In the case of mandarin, TW-RDI showed the highest decline 342
(79 and 60% for A and gs, respectively).
343
Leaf structural traits: leaf dry mass, phytotoxic elements and chlorophyll.
344
LMA is positively related to leaf photosynthetic capacity (Niinemets, 1999), hence 345
grapefruit trees presented higher values of LMA than mandarin trees (Table 5), as 346
expected from gas exchange measurements. There were also significant differences 347
between treatments: the highest LMA values were observed in TW treatments for 348
grapefruit trees and in RW-RDI for mandarin (Table 5).
349
Regarding phytotoxic elements (Table 5), RW-C treatment showed Cl-, Na and B levels 350
significantly higher than TW treatments in both crops, except to the B in mandarin. In 351
agreement with the phytotoxic thresholds reported by Romero-Trigueros et al. (2014b), 352
in our study the Na limit was not exceeded by any treatment, Cl- only by RW-C of 353
mandarin and B by both RW treatments on grapefruit and RW-RDI on mandarin.
354
Moreover, differences in leaf chlorophyll content can be an indicator of photosynthetic 355
capacity and degree of stress (Wu et al., 2008). In addition, the coefficient Chl a/Chl b 356
(Coef a/b) can be used as an index to characterize the plant physiological status. In our 357
study, RW treatments of both crops showed the lowest values of total chlorophyll, Chl T 358
(Figure 3) and the highest values of Coef a/b, in accordance with Bondada and 359
Syvertsen (2003). Only in RW treatments of mandarin the Coef a/b increased from t1 to 360
t2 (Figure 3C and 3D) due to a decrease in Chl b since increments in radiance destroy 361
the Chl b in greater proportion than Chl a due to the fact that photosystem II, which is 362
rich in Chl b, becomes more unstable (Casierra-Posada, 2007).
363
3.2 Spectral indicators in Citrus species 364
In general, we observed that reflectance in the NIR region was about 7% higher in 365
Control grapefruit than in Control mandarin trees whereas the reflectance values in the 366
R wavelength were about 3% lower in control grapefruit than in Control mandarin trees 367
at t1. No differences were detected at t2 between species. It is noticeable that R and NIR 368
reflectance decreased from t1 to t2 within all mandarin and grapefruit treatments due to 369
changes in climatic conditions (solar radiation, air temperature, VPD, etc.).
370
Grapefruit 371
At t1, trees under water and salt stress (TW-RDI, RW-C and RW-RDI) showed a 372
significant increase in the reflectance on the R domain with respect to TW-C (Table 373
6A). This isin contrast with what Contreras et al. (2014) found for the same plot at the 374
beginning of the RW application in 2009. This increase in R responds to the observed 375
decrease in Chl T in those treatments (Figure 3A). On the contrary, no significant 376
differences between treatments were found in the NIR region. The NDVI was 377
significantly higher in TW than RW treatments (Table 6A). Similar results were 378
obtained by Contreras et al. (2014). At t2, only trees irrigated with RW showed an 379
increase in the R domain, coinciding again with Chl T (Figure 3A). NIR reflectance in 380
this second AF was significantly lower in both RDI treatments (TW-RDI and RW-RDI) 381
but not in RW-C (Table 6A), in accordance with lower Ψs levels (Figure 2A).
382
Mandarin 383
At t1, the highest R values were observed in RW treatments. The RW-RDI had the 384
biggest effect, probably as a result of the low chlorophyll concentration (Figure 3B).
385
Regarding the NIR region, trees under deficit irrigation (RDI treatments) had higher 386
values than C trees, in accordance with Ψs data (Figure 2B). Moreover, in contrast to 387
grapefruit, the trees with significantly higher NDVI values were those in the C 388
treatments, regardless of water quality. At t2, R increased only with TW-RDI (Table 6B) 389
and not with RW treatments also, as expected it would do in relation to chlorophyll 390
decreases (Figure 3B).
391
It is thus worth highlighting that the ΨP increase in RW treatments (Table 3), due to a 392
low Ψπ driven by Cl- and Na from RW, likely interfered with R reflectance. Finally, 393
there were no significant differences among treatments for NIR.
394
3.3 Correlations between spectral indicators and plant water status and leaf 395
structural traits.
396
Red domain (R) 397
On grapefruit trees (Table 7A), the R domain was significantly correlated with Chl T 398
and Coef a/b (p<0.01 and p<0.05, respectively) as expected according to the data shown 399
in sections 3.1 and 3.2. This correlation was negative since R reflectance is lower with 400
increasing chlorophyll. Sims and Gamon (2002) and Ollinger (2011) demonstrated that 401
the R domain was linked to the photosynthetic leaf pigments across a wide range of 402
species. Because of important physiological roles of leaf chlorophyll and its strong 403
absorbance properties, it is important have corroborated that the method here evaluate 404
using UAVs is a useful and effective tool to estimate Chl T from grapefruit canopy 405
reflectance and that avoids destructive laboratory methods. Moreover, the R domain 406
was also significantly linked to ΨP. This was associated to the fact that absorbance 407
includes light absorbed by pigments, as we observed with R absorbance by Chl T, but 408
maybe also by other leaf constituents (Kokaly et al., 2009)such as those associated with 409
the increased turgor.
410
On mandarin trees, the R domain was significantly related to Ψs, A and gs according to 411
Sims and Gamon (2002). To the contrary, no significant correlation between the R and 412
Chl T was observed since the R values found in the RW treatments were lower than 413
expected, as the Chl T concentration at t2 (Figure 3B). Consequently, under high VPD 414
conditions reflectance of mandarin trees (at t2) was stronger influenced by gas 415
exchange, Ψπ and ΨP than by chlorophyll (RW treatments showed the highest Ψπ and 416
ΨP, Table 3).
417
Near infrared domain (NIR) 418
The biophysical basis for high leaf-level reflectance in the NIR region is provided by 419
(Ollinger, 2011). It is related to the likelihood of photons being scattered from the point 420
of entry into the leaf because absorption by leaf constituents is either small or altogether 421
absent (Merzlyak et al., 2002). In our study, NIR for both grapefruit and mandarin trees 422
was positively linked to Ψs and consequently with gas exchange parameters, as we 423
expectedfrom the results of sections 3.1 and 3.2. High values of net photosynthesis (A) 424
correlated with high NIR values, likely as a result of scattering in the NIR region caused 425
by high CO2 levels in leaves (Ollinger, 2011).
426
NDVI index 427
The NDVI index for grapefruit trees had a direct relationship with A and gs in 428
accordance with data reported by Baluja et al. (2012) and Gago et al. (2015) for 429
vineyards, and Zarco-Tejada et al. (2012) for Citrus. The NDVI for mandarin trees 430
correlated well with Ψs, in agreement with the findings of Baluja et al. (2012). NDVI 431
and other vegetation indices proposed to monitor vegetation dynamics are considered 432
structural indices related to plant vigor (Dobrowski et al., 2005; Gago et al., 2015;
433
González-Dugo et al., 2015; Zarco-Tejada et al., 2013b) as they track changes in canopy 434
structure but have little or no sensitivity to short-term leaf physiological changes which 435
are independent of canopy structure according to Haboudane et al. (2004). However, the 436
current work showed that in case of Citrus, NDVI responds to short-term changes in gas 437
exchange and Ψs. Thus, we can confirm that NDVI can be sensitive in Citrus to diurnal 438
physiological changes induced by variations in environmental conditions throughout the 439
day and not only tracks the effects in the long term as other authors indicated 440
(Dobrowski et al., 2005; Zarco-Tejada et al., 2013c). Similar conclusions were obtained 441
Baluja et al. (2012) for vineyard crop.
442
Best indicators across species 443
Bearing in mind data from both species together (Figure 4), NIR was significantly 444
correlated with Ψs (p<0.005) and R with Chl T (p<0.005). For the last one, it was 445
necessary to eliminate the point from the RW treatment at t2 of mandarin due to –as was 446
mentioned above- when mandarin trees were under high values of VPD (at t2), the R 447
domain is more influenced by gas exchange, Ψπand ΨP, than by chlorophyll. Therefore, 448
we considered the NIR and R spectral indicators as the best related to the parameters 449
measured at the leaf scale for Citrus crops.
450
4. Conclusions 451
This study assessed the effects of eight years of irrigation with RW and deficit irrigation 452
on grapefruit and mandarin trees on a diurnal basis. The results suggest that on 453
grapefruit trees the water potential was affected by water stress (RDI) but not by saline 454
stress when trees were well irrigated with RW. Gas exchange was reduced by both 455
stresses. The water potential of mandarin trees was not affected by any treatment and 456
gas exchange was only reduced by RDI with TW. The total chlorophyll of both crops 457
decreased with RW treatments.
458
Regarding spectral data, for grapefruit, R wavelength values increased with RW 459
treatments, consistent with chlorophyll data, and the NDVI levels decreased at 07.00 460
GMT since gas exchange also declined. The NIR region was affected mainly by deficit 461
irrigation, regardless water quality, in the second airborne flight. For mandarin, R 462
domain increased with declining of chlorophyll in RW treatments. However, when 463
climatic conditions were more stressful, R was influenced mainly by the increasing leaf 464
turgor and gas exchange. Therefore, the response in R was attributed to stress-induced 465
declines in leaf chlorophyll. But when VPD was too high, R could detect physiological 466
changes in other parameters and responded in a shorter term than those related 467
exclusively with the chlorophyll synthesis. NIR was linked to deficit irrigation 468
treatments and NDVI only increased under well irrigated conditions, regardless of water 469
quality.
470
Because all of the above, we obtained significant correlations between: i) For grapefruit:
471
R with chlorophyll and potential turgor; NIR with Ψs and gas exchange (A and gs); and 472
NDVI with gas exchange. ii) For mandarin: R correlated with chlorophyll only at the 473
first hour of the morning; NIR with stem water potential and gas exchange, as in 474
grapefruit, and NDVI with stem water potential.
475
We conclude the following: The statistical analyses of field data and remote sensing 476
data, derived from multispectral imagery using an UAV, confirms the feasibility of 477
applying the proposed methods to assess physiological and structural properties of 478
Citrus under water and saline stress.
479 480
Acknowledgment 481
This study was supported by two CICYT (AGL2010-17553 and AGL2013-49047-C2- 482
515 2-R) projects and SIRRIMED (KBBE-2009-1-2-03, PROPOSAL N◦245159) 483
project. We are also grateful to SENECA–Excelencia Científica (19903/GERM/15) for 484
providing funds for this research.
485 486
References 487
Aksoy, U., Hepaksoy, S., Can, H.Z., Anaç, S., Ul, M.A., Dorsan, F., Anaç, D., Okur, B., 488
Kiliç, C., 1998. The effect of rootstock on leaf characteristics and physiological 489
response of satsuma mandarins under saline conditions. Acta Hortic. 513, 169-176. doi:
490
10.17660/ActaHortic.1998.513.19.http://dx.doi.org/10.17660/ActaHortic.1998.513.19 491
Allen, R.G., Pereira, L.S., Raes, D., Smith, M., 1998. Crop evapotranspiration- 492
guidelines for computing crop water requirements. FAO Irrig. Drain. 56, 15-27.
493
Anderson, K., Gaston, K.J., 2013. Lightweight unmanned aerial vehicles will 494
revolutionize spatial ecology. Front. Ecol. Environ. 11(3), 138–146.
495
Anjum, M.A., 2008. Effect of NaCl concentrations in irrigation water on growth and 496
polyamine metabolism in two citrus rootstocks with different levels of salinity 497
tolerance. Acta Physiol. Plant. 30, 43-52.
498
Baluja, J., Diago, M.P., Balda, P., Zorer, R., Meggio, F., Morales, F., Tardaguila, J., 499
2012. Assessment of vineyard water status variability by thermal and multispectral 500
imagery using an unmanned aerial vehicle (UAV). Irrig. Sci. 30(6), 511-522.
501
Berni, J.A.J., Zarco-Tejada, P.J., Sepulcre-Cantó, G., Fereres, E., Villalobos, F., 2009a.
502
Mapping canopy conductance and CWSI in olive orchards using high resolution thermal 503
remote sensing imagery. Remote Sens. Environ. 113(11), 2380-2388.
504
Berni, J.A.J., Zarco-Tejada, P.J., Suárez, L., Fereres, E., 2009b. Thermal and narrow 505
band multispectral remote sensing for vegetation monitoring from an unmanned aerial 506
vehicle. Geosci. Remote Sens. IEEE Trans. 47(3), 722–738.
507
http://dx.doi.org/10.1109/TGRS.2008.2010457.
508
Bondada, B.R., and Syvertsen, J.P., 2003. Leaf chlorophyll, net gas exchange and 509
chloroplast ultrastructure in citrus leaves of different nitrogen status. Tree Physiol.
510
23(8), 553-559.
511
Bonilla, I., Martínez, F., Martínez-Casasnovas, J.A., 2015. Vine vigor, yield and grape 512
quality assessment by airborne remote sensing over three years: Analysis of unexpected 513
relationships in cv. Tempranillo. Span. J. Agric. Res. 13(2), e0903, 8 pages. eISSN:
514
2171-9292. http://dx.doi.org/10.5424/sjar/2015132-7809.
515
Casierra-Posada, F., Ávila-León, O.F., Riascos-Ortíz, D.H., 2012. Diurnal changes in 516
photosynthetic pigments content in sun and shade marigold leaves. Temas Agrarios. 17, 517
60- 71.Chalmers, D.J., Mitchell, P.D., Van Heek, L., 1981. Control of peach tree growth 518
and productivity by regulated water supply, tree density and summer pruning. J. Am.
519
Soc. Hort. Sci. 106, 307–312.
520
Contreras, S., Pérez-Cutillas, P., Santoni, C.S., Romero-Trigueros, C., Pedrero, F., 521
Alarcón, J.J., 2014. Effets of reclaimed waters of spectral properties and leaf traits of 522
citrus orchards. Water Environ. Res. 86, 2242-2250.
523
Desta, K.G., 2014. Washing Plant Tissue Samples for Mineral Nutrient Analysis.
524
Washington State University, USA. Extension fact sheet FS134E.
525
Dobrowski, S.Z., Pusknik, J.C., Zarco-Tejada, P.J., Ustin, S.L., 2005. Simple 526
reflectance indices track heat and water stress induced changes in steady state 527
chlorophyll fluorescence. Remote Sens. Environ. 97(3), 403-414.
528
Fereres, E., Soriano, M., 2007. Deficit irrigation for reducing agricultural water use. J.
529
Exp. Bot. 58, 147-159.
530
Flexas, J., Bota, J., Escalona, J.M., Sampol, B., Medrano, H., 2002. Effects of drought 531
on photosynthesis and electron transport rate regulation in grapevine. Plant Cell 532
Environ. 22, 39-48.
533
Gago, J., Martorell, S., Tomás, M., Pou, A., Millán, B., Ramón, J., Ruiz, M., Sánchez, 534
R., Galmés, J., Conesa, M.A., Cuxart, J., Tardáguila, J., Ribas-Carbó, M., Flexas, J., 535
Medrano, H., Escalona, J.M., 2013. High-resolution aerial thermal imagery for plant 536
water status assessment in vineyards using a multicopter-RPAS. In: "VII Congreso 537
Ibérico de Agroingeniería y Ciencias Hortícolas. Sociedades Españolas de 538
Agroingeniería y de Ciencias Hortícolas, y las Sociedades Portuguesas de Horticultura y 539
la Sección Especializada de Ingeniería Rural de la Sociedad de Ciencias Agrarias de 540
Portugal.", Madrid, España. Poster.
541
Gago, J., Douthe, C., Coopman, R.E., Gallego, P.P., Ribas-Carbo, M., Flexas, J., 542
Escalona, J., Medrano, H., 2015. UAVs challenge to assess water stress for sustainable 543
agriculture. Agric. Water Manage. 153, 9-19.
544
Garcia-Galiano, S.G., Olmos-Giménez, P., Giraldo-Osorio, J.D., 2015. Assessing 545
Nonstationary Spatial Patterns of Extreme Droughts from Long-Term High-Resolution 546
Observational Dataset on a Semiarid Basin (Spain). Water. 7(10), 5458-5473.
547
doi:10.3390/w7105458 548
García-Sánchez, F., Syvertsen, J.P., 2006. Salinity reduces growth, gas exchange, 549
chlorophyll and nutrient concentrations in diploid sour orange and related allotetraploid 550
somatic hybrids. J. Hort. Sci. Biotechnol. 77, 379-386.
551
García-Tejero, I., Jiménez-Bocanegra, J.A., Martínez, G., Romero, R., Durán-Zuazo, 552
V,H., Muriel-Fernández, J.L., 2010. Positive impact of regulated deficit irrigation on 553
yield and fruit quality in a commercial citrus orchard. Agric. Water Manage. 97, 614- 554
622.
555
Geoghegan-Quin, M., 2013. Role of Research & Innovation in Agriculture. European 556
Commission-SPEECH/13/505. http://europa.eu/rapid/press-releaseSPEECH-13-505 557
en.htm 558
Gimeno, V., Syvertsen, J. P., Nieves, M., Simón, I., Martínez, V., Garcia-Sanchez, F., 559
2009. Orange varieties as interstocks increase the salt tolerance of lemon trees. J.
560
Hortic. Sci. Biotechnol. 84(6), 625-631. doi: 10.1080/14620316.2009.11512577.
561
Gimeno, V., Simón, I., Nieves, M., Martínez, V., Cámara-Zapata, J.M., García, A.L., 562
García-Sánchez, F., 2012. The physiological and nutritional responses to an excess of 563
boron by Verna lemon trees that were grafted on four contrasting rootstocks. Trees- 564
Struc. Funct. 26(5), 1513-1526.
565
Glenn, E.P., Huete, A.R., Nagler, P.L., Nelson, S.G., 2008. Relationship between 566
Remotely-Sensed Vegetation Indices, Canopy Attributes and Plant Physiological 567
Processes: What Vegetation Indices Can and Cannot Tell Us about the Landscape.
568
Sensors. 8, 2136–2160.
569
Gomes, M.M., Lagoa, A.M.M.A., Medina, C.L., Machado, E.C., Machado, M.A., 2004.
570
Interactions between leaf water potential, stomatal conductance and abscisicacid content 571
of orange trees submitted to drought stress. Braz. J. Plant. Physiol. 13(3), 155–161.
572
González-Dugo, V., Zarco-Tejada, P., Berni, J.A., Suárez, L., Goldhamer, D., Fereres, 573
E., 2012. Almond tree canopy temperature reveals intra-crown variability that is water 574
stress-dependent. Agric. Forest. Meteorol. 154, 156–165.
575
González-Dugo, V., Zarco-Tejada, P., Nicolás, E., Nortes, P.A., Alarcón, J.J., 576
Intrigliolo, D.S., Fereres, E., 2013. Using high resolution UAV thermal imagery to 577
assess the variability in the water status of five fruit tree species within a commercial 578
orchard. Precis. Agric. 14(6), 660–678.
579
Gonzalez-Dugo, V., Hernandez, P., Solis, I., Zarco-Tejada, P.J., 2015. Using High- 580
Resolution Hyperspectral and Thermal Airborne Imagery to Assess Physiological 581
Condition in the Context of Wheat Phenotyping. Remote Sens. 7, 13586-13605.
582
Gucci, R., Xiloyannis, C., Flore, J.A., 1991. Gas exchange parameters water relations 583
and carbohydrate partitioning in leaves of field-grown Prunus domestica following fruit 584
removal. Physiol. Plant. 83, 497–505.
585
Haboudane, D., Miller, J.R., Pattey, E., Zarco-Tejada, P.J., Strachan, I., 2004.
586
Hyperspectral vegetation indices and novel algorithms for predicting green LAI of crop 587
canopies: Modeling and validation in the context of precision agriculture. Remote Sens.
588
Environ. 90(3), 337-352.
589
Hussain, S., Luro, F., Costantino, G., Ollitrault, P., Morillon, R., 2012. Physiological 590
analysis of salt stress behavior of citrus species and genera: Low chloride accumulation 591
as an indicator of salt tolerance. South Afr. J. Bot. 81, 103-112.
592
Jones, H.G., Vaughan, R.A., 2010. Remote Sensing of Vegetation: Principles, 593
Techniques and Applications. Oxford University Press Inc., New York.
594
Kokaly, R.F., Asner, G.P., Ollinger, S.V., Martin, M.E., Wessman, C.A., 2009.
595
Characterizing canopy biochemistry from imaging spectroscopy and its application to 596
ecosystem studies. Remote Sens Environ. 113, 78-91. doi: 10.1016/j.rse.2008.10.018.
597
Lelong, C.C., Burger, P., Jubelin, G., Roux, B., Labbé, S., Baret, F., 2008. Assessment 598
of unmanned aerial vehicles imagery for quantitative monitoring of wheat crop in small 599
plots. Sensor. 8(5), 3557-3585. http://dx.doi.org/10.3390/s8053557.
600
Levy, Y., Syvertsen, J.P., 2004. Irrigation water quality and salinity effects in citrus 601
trees. Hortic. Rev. 30, 37-82.
602
Lucieer, S.M., Jong, D., Turner, D., 2013. Mapping landslide displacements using 603
Structure from Motion (SfM) and image correlation of multi-temporal UAV 604
photography, Prog. Phys. Geogr. 38(1), 97–116. doi:10.1177/0309133313515293.
605
Melgar, J.C., 2008. Leaf gas exchange, water relations, nutrient content and growth in 606
citrus and olive seedlings under salinity. Biol. Plant. 52, 385-390.
607
Merzlyak, M.N., Chivkunova, O.B., Melo, T.B., Naqvi, K.R., 2002.Does a leaf absorb 608
radiation in the near infrared (780-900 nm) region? A new approach to quantifying 609
optical reflection, absorption and transmission of leaves. Photosyn. Res. 72(3), 263-270.
610
Navarro, J.M., García-Olmos, B., Andujar, S., Rodríguez-Morán, M., Moreno, M., 611
Porras, I., 2011. Effects of calcium on growth and nutritional state of citrus seedlings 612
under NaCl stress. Acta Hortic. 922, 55-60.
613
Nicolás, E., Alarcón, J.J, Mounzer, O., Pedrero, F., Nortes, P.A., Alcobendas, R., 614
Romero-Trigueros, C., Bayona, J.M., Maestre-Valero, J.F., 2016. Long-term 615
physiological and agronomic responses of mandarin trees to irrigation with saline 616
reclaimed water. Agric. Water Manage. 166, 1-8.
617
Niinemets, U., 1999. Research review. Components of leaf dry mass per area-thickness 618
and density-alter leaf photosynthetic capacity in reverse directions in woody plants.
619
New Phytol. 144, 35-47.
620
Ollinger, S.V., 2011. Sources of Variability in Canopy Reflectance and the Convergent 621
Properties of Plants. New Phytol. 189, 375–394.
622
Papadakis, I.E., Dimassi, K.N., Bosabalidis, A.M., Therios, I.N., Giannakoula, A., 623
2004. Effects of B excess on some physiological and anatomical parameters of 624
‘Navelina’ orange plants grafted on two rootstocks. Environ. Exp. Bot. 51, 247-257.
625
Paranychianakis, N.V., Chartzoulakis, K.S., Angelakis, A.N., 2004. Influence 626
ofrootstock, irrigation level and recycled water on water relations and leaf gas exchange 627
of Soultanina grapevines. Environ. Exp. Bot. 52, 185–198.
628
Pedrero, F., Mounzer, O., Alarcón, J.J., Bayona, J.M., Nicolás, E., 2013. The viability 629
of irrigating mandarin trees with saline reclaimed water in a semi-arid mediterranean 630
region: a preliminary assessment. Irrig. Sci. 31(4), 759-768.
631
Pedrero, F., Maestre-Valero, J.F., Mounzer, O., Nortes, P.A., Alcobendas, R., Romero- 632
Trigueros, R., Bayona, J.M., Alarcón, J.J., Nicolás, E., 2015. Response of young ‘Star 633
Ruby’ grapefruit trees to regulated deficit irrigation with saline reclaimed water. Agric.
634
Water Manage. 158, 51-60.
635
Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the Köppen- 636
Geiger climate classification. Hydrol. Earth Syst. Sc. 11, 1633–1644.
637
Pérez-Pérez, J.G., Syvertsen, J.P., Botía, P., García-Sánchez, F., 2007. Leaf Water 638
Relations and Net Gas Exchange Responses of Salinized Carrizo Citrange Seedlings 639
during Drought Stress and Recovery. Ann. Bot. 100(2), 335-345.
640
Pôcas, I., Rodrigues, A., Gonçalvez, S., Costa, P.M., Gonçalves, I., Pereira, L.S., 641
Cunha, M., 2015. Predicting grapevine water status based on hyperspectral reflectance 642
vegetation indices. Remote Sens. 7, 16460-16479. doi: 10.3390/rs71215835.
643
Rodriguez-Pérez, J.R., Riaño, D., Carlisle, E., Ustin, S., Smart, D.R., 2007. Evaluation 644
of hyperspectral reflectance indexes to detect grapevine water status in vineyards. Am.
645
J. Enol. Vitic. 58, 302-317.
646
Romero-Trigueros, C., Nortes, P.A., Alarcón. J.J., Nicolás, E., 2014a. Determination of 647
15N stable isotope natural abundances for assessing the use of saline reclaimed water in 648
grapefruit. Environ. Eng. Manag. J. 13(10), 2525-2530.
649
Romero-Trigueros, C., Nortes, P.A., Pedrero, F., Mounzer, O., Alarcón, J.J., Bayona, 650
J.M., Nicolás, E., 2014b. Assessment of the sustainability of using saline reclaimed 651
water in grapefruit in medium to long term. Span. J. Agric. Res. 12(4), 1137-1148.
652
Scholander, P.F., Hammel, H.T., Bradstreet, E.D., Hemmingsen, E.A., 1965. Sap 653
pressure in vascular plants. Science. 148, 339-434 654
Shackel, K., Ahmadi, H., Biasi, W., Buchner, R., Goldhamer, D., Gurusinghe, S., 655
Hasey, J., Kester, D., Krueger, B., Lampinen, B., McGourty, G., Micke, W., Mitcham, 656
E., Olson, B., Pelletrau, K., Philips, H. Ramos, D., Schwankl, L., Sibbett, S., Snyder, R., 657
Southwick, S., Stevenson, M., Thorpe, M., Weinbaum, S., Yeager, J., 1997. Plant water 658
status as an index of irrigation need in deciduous fruit trees. Hort. Technol. 7(1), 23–29.
659
Sims, D.A., Gamon, J.A., 2002. Relationships between leaf pigment content and 660
spectral reflectance across a wide range of Species, Leaf structures and developmental 661
stages. Remote Sens. Environ. 81, 337-354.
662
